In the field of video compression, communication, decompression, and display, there have been for many years problems associated with supporting both interlaced content and interlaced displays along with progressive content and progressive displays. There are serious technical issues with merging progressive systems and interlaced systems, and as a result there are separate, largely incompatible systems in use. There are strong motivations to design a common system, to have content-compatible systems for all types of displays, to support the coming generation of progressive display Televisions (TVs), and to support the most efficient compression of video in new compression standards.
Digital compression schemes that are optimized for interlaced video content are more complex than those designed for progressive video, and the degree of added complexity associated with interlaced coding increases as other aspects of the compression schemes also become more complex, in the ongoing effort to provide more efficient compression. As a result, the use of interlaced coding tools is becoming increasingly less attractive.
Conventional TVs, and almost all TVs in use today, use the interlaced scanning method that was introduced many decades ago. Virtually all TV broadcasts and recorded video content intended for TV display are in interlaced format. On the other hand, almost all personal computers (PCs) use progressive scanning displays, and compressed video available over the Internet and compressed video for PCs are in progressive format as well. Hence, there is a problem now in that TVs use one format, while PCs use another.
In the realm of high definition televisions (HDTV), there was a conflict whether interlaced format or progressive format should be used for broadcast, and the result is that both formats are in use, with the requirement that HDTVs be able to convert between them to support both types of content and, in some cases, both types of display.
A set of HDTV video formats, as defined by Advanced Television Standards Committee (ATSC), includes a progressive format at 1280×720/60p (60 frames per second (fps) progressive) and an interlaced format 1920×1080/30i (30 fps interlaced). These two formats are commonly used, and there are converters that convert back and forth between them. However, the standard for the progressive format of 1280×720 only supports frames or images with 1280 pixels per line or less, which requires the interlaced format of 1920×1080/30i to be horizontally scaled down to 1280 or less, which may not be desirable. This standard also requires using the interlaced coding tools, which increase the complexity of a system.
The problem many systems are faced with is the complexity of digitally compressing interlaced content. Fields of interlaced frames are spatially and temporally separated. One approach to deal with this problem is to treat interlaced video frames as if each field were a progressive video frame. There are several problems with this approach. Each field contains only half the vertical resolution of the original frame, so treating a field as if it were a frame causes a loss of resolution and aliasing of vertical detail. Also, the even and odd fields represent different vertical sampling positions, making motion-compensated compression more difficult.
Another approach that has been used is to take 30 fps interlaced video, which is 60 fields per second, and convert it to 30 fps progressive by some means. One problem in such an approach is that temporal detail is being thrown away, because each field of interlaced video is sampled at a unique time, with 60 samples per second in this example, and the result of this conversion has only 30 samples per second. Methods that convert interlaced video into progressive video format, known as de-interlacers, are well known in the art. A representative published paper that gives an overview of de-interlacing methods is: “Deinterlacing-an overview,” De Haan, G.; Bellers, E. B.; Proceedings of the EEE, Volume: 86 Issue: 9, September 1998 Page(s): 1839-1857.
Other compression techniques have been used by standards such as the MPEG-2 standard where interlaced video is directly compressed. One method uses frame pictures where two interlaced fields are interleaved directly and treated as a frame. Another method uses field pictures where each interlaced field is treated as a picture. Yet another method uses field coded frame pictures where frame pictures resulting from interleaving two interlaced fields are compressed using static or dynamic field and frame coding methods, including inter-frame prediction using frame or field prediction and field or frame transform coding of the residual signal.
The MPEG AVC (Audio-Visual Coding) standard (a.k.a. ITU H.264) is another example of a standard that uses a compression method that includes explicit coding tools for interlaced content. In the example of MPEG AVC, the combination of explicit interlaced coding with an adaptive loop filter, direct mode motion compensation and other modern compression techniques results in substantially increased complexity of both encoders and decoders.
In order to make widespread use of new compression formats, it is highly desirable that the vast range of content meant for TVs be supported, and that TV displays also be supported, in an efficient and cost effective manner, without adding undue complexity to the compression and decompression systems.
Further limitations and disadvantages of conventional and traditional approaches will become apparent to one of ordinary skill in the art through comparison of such systems with the present invention.